[Development fund for Africa --] impact evaluation of the Liberia Lofa County agricultural development project

Project Phases I and II were implemented between 1976 and 1987. The Government of Liberia has been progressively retiring or absorbing the project infrastructure and staff since 6/87, and the Ministry of Agriculture was to have finished this process by the end of 1988. LCADP has had a profound impact on small farmers in Upper Lofa. Over 10,000 farmers and their households benefitted from improvements in rural infrastructure and agricultural inputs. Smallholder acreage of cocoa and coffee has increased dramatically, and improved varieties of upland and swamp rice are now widely used. LCADP built many roads, offered health services, and improved village water supplies with hundreds of wells. The project focused on promoting new farming techniques and service delivery to the rural communities. The Project Management Unit -- created as an institution separate from the weak Ministry of Agriculture -- should have ultimately transferred services delivery and extension work to the regional cooperatives, but this approach was based on a false assessment of these cooperatives. The cooperatives have failed to become independent multi-functional institutions. The size and basis for recruitment (farmers were compelled to join to receive LCADP technical inputs) robbed even pre-existing community associations of any genuine popular participation. The project also failed to furnish the training and guidance needed to generate viable institutions. The centralized, hierarchical administrative structure of the cooperatives did not correspond to local social organizations, patterns of authority, and small farmer interests and aspirations. The cooperatives failed even in their most basic role of marketing their members' coffee and cocoa, although the blame is not entirely theirs. The credit component of LCADP was not central to design or implementation. The project delivered farming inputs and associated credit based on targets, not on assessments of credit risks and likely recovery. Farmer loan repayment was intended to create a revolving fund to sustain future cooperative lending. However, LCADP never established the management systems necessary to follow the loan portfolio properly. The low recovery rate of only 0.5 percent on the long-term "development" loans illustrates the failure of the credit component. External factors compounded the impact of the original weaknesses in project formulation and execution. Negative national economic growth, sharp declines in world prices for cocoa and coffee, periodic illiquidity of institutions key to farmers (such as the Agricultural and Cooperative Development Bank and the Liberian Produce Marketing Corporation), and other factors all contributed to the failure of the credit component as well. The evaluation's major findings are as follows: (1) The model of an area development project strengthening and developing existing cooperatives to assume the lead in agricultural development and farmer credit failed. (2) Based on repayment rates, small-farmer lending by the LCADP Project Management Unit did not work. (3) The model of an agricultural development project building up a sustainable revolving credit fund for meeting future small-farmer credit needs failed. The LCADP fund proved difficult to manage as intended, and, in any case, probably represented an overestimation of the need for externally supplied credit. (4) The model of tying credit to particular input packages failed in the sense that it did not assure precise adoption of recommended practices, nor render credit an infungible farming input. (5) Tying credit to a limited number of cropping packages failed to capitalize on other productive opportunities, especially for women. (6) The external economic and political environment is extremely important to the functioning of rural credit projects. (7) Since it lent without great regard to orthodox creditworthiness criteria, such as preexisting collateral, the LCADP was able to reach poorer farmers with its loans. (8) There are a variety of formal and informal deposit-taking and lending institutions in Upper Lofa County that are widely used by small farmers. The lessons learned based on the above are: (1) Developing true farmer-serving, multifunctional cooperatives is difficult. At least initially, farmer associations should be small community groups based on local participation and leadership. Such groups should be analyzed to determine their capacity to serve as the basis for effective, new organizations such as marketing groups or credit unions -- before attempts are made to establish these larger, more complex entities. (2) Small-farmer lending by multifunctional cooperatives may conflict with their other roles, such as marketing and TA and may be misunderstood by farmers. Wiser is spinning off by farmer cooperatives, or separate establishment, of fully bank-like, deposit-taking credit cooperatives. (3) Giving a marketing cooperative a monopoly over procurement from its members disinclines it to serve its members as well as possible. The Liberia findings suggest that free wholesale-market competition is a necessary spur to insuring that cooperatives exercise their inherent size and ownership advantages to serve all their members well, not just a dominant few. (4) Small-farmer lending by agricultural development project management units, or any other entity that is not a true financial institution, should be avoided. The Liberia case confirms the view that farmer-borrowers at least partly condition repayment on their desire for access to future credit from, and stake in (i.e., deposits), the lending institution. (5) Donors should think twice about undertaking major agricultural development projects where dealing with a monopoly-holding marketing parastatal cannot be avoided. In the past decade, Liberia's coffee and cocoa parastatal has served Lofa County farmers poorly, ultimately nullifying many of the project's gains. (6) Revolving credit funds of specially earmarked funds to service a region's future agricultural borrowing needs are unlikely to work -- particularly when the monies must flow through a number of hands -- and are best avoided. (7) Tying credit to particular input packages is unnecessary. The Liberia study confirmed findings elsewhere that this approach does not ensure precise adoption of recommended practices, nor render production credit infungible. Institutional credit for inputs is better given in cash. (Author abstract)